Identification of Cellular Damage in Uteri From Hyperinsulinemic Mice Treated With Unopposed Estradiol

Abstract Unopposed estradiol and obesity are known risk factors for endometrial adenocarcinoma (EC). Endometrium from women with obesity was found to have an increase in mutations relative to tissue from normal weight women, indicating DNA damage may be accelerated in the setting of obesity. Since obesity is associated with high levels of insulin and anovulatory cycles, we sought to mimic these conditions in a mouse model. We previously found that hyperinsulinemic MKR mice, without the confounder of obesity, have an increased incidence of nuclear atypia in endometrial glands. We hypothesized that hyperinsulinemia and unopposed estradiol have a synergistic effect on inducing abnormal architecture and DNA damage in the endometrium, than either alone. At 8-10 weeks old, cohorts of MKR (n=20) and WT (n=20) mice underwent ovariectomy and placement of either an estradiol (E2) or placebo (P) pellet. Metabolic profiling included insulin tolerance testing and MR for body composition. At 3 months post-implantation, mice received a partial hysterectomy and second pellet replacement. At 6 months, the remaining uterus was bisected into pieces. A blinded histological analysis was conducted by a gynecology pathologist. A marker of DNA damage due to oxidative stress, 8-oxoguanine-DNA-glycosylase (8-OHdG), was quantified by ELISA. Data was analyzed using Kruskal-Wallis test with multiple test correction, or Fischer’s exact test. By 6 months, MKR-E2 treated mice had a 27% lower body weight than MKR-P mice (p<0.05), and 31% lower than WT-E2 mice (p<0.01). WT-E2 and WT-P had similar weight, and were similar to MKR-P (p=ns). Percent body fat was similar across all 4 cohorts of mice (p=ns). Since placebo-treated mice had small, atrophied uteri with minimal gland formation, E2 pellet failure was determined by the presence of small, atrophied uteri and occurred in 4 MKR and 3 WT mice at either 3 or 6 months. All other MKR and WT E2 treated mice had enlarged uteri. The frequency of endometrial gland dilation was similar in MKR-E2 and WT-E2 uteri (p=ns), but all MKR mice had moderate-severe dilation, whereas WT mice had 50% mild and 50% moderate-severe dilation (p=0.07). Focal hyperplasia was present in one MKR-E2 mouse, and nuclear atypia was present in one WT-E2 mouse. MKR-P uteri had a 7-fold higher mean 8-OHdG relative to MKR-E2 uteri (5.0±3.7 vs 0.7±1.6, p<0.002). WT-E2 and WT-P uteri had similar 8-OHdG (1.6±0.8 vs 1.8±0.6, p=ns), as did MKR-E2 and WT-E2 uteri (p=ns). Our findings show that hyperinsulinemia exacerbates the cystic dilation induced by chronic unopposed estradiol, indicating a synergy of insulin and estradiol in promoting abnormal glandular growth in the endometrium. Surprisingly, uterine DNA damage was highest in the setting of hyperinsulinemia alone, in a hormonal state mimicking post-menopause. Further work is needed to understand the effect of estradiol on intrauterine oxidative stress-induced damage.